Saurabh Jaiswal

Key Skills

Software

Clickworker

Data Annotation Tech

Mercor

Mindrift

CVAT

Top Subject Matter

Large Language Models (LLMs)

Nlp Domain Expertise

Conversational AI

Top Data Types

3D Sensor

Audio

Computer Code Programming

Document

Geospatial Tiled Imagery

Image

Medical Dicom

Text

Video

Top Task Types

Computer Programming/Coding

Data Collection

Prompt + Response Writing (SFT)

Text Generation

Classification

Bounding Box

Freelancer Overview

AI Model Training and Evaluation Annotator. Core strengths include Mindrift, CVAT, and N. AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

Refinery Image Dataset Validation and Label Optimization

CVATImageBounding Box

Validated and optimized auto-generated bounding boxes and class labels for refinery image data. Analyzed annotation quality and enhanced labeling consistency in highly complex visual scenes. Ensured accuracy for building computer vision model training sets. • Detected labeling errors in auto-annotated images • Improved precision of bounding box localization • Enhanced classification granularity for multiple categories • Supported annotation best practices for industrial datasets

Validated and optimized auto-generated bounding boxes and class labels for refinery image data. Analyzed annotation quality and enhanced labeling consistency in highly complex visual scenes. Ensured accuracy for building computer vision model training sets. • Detected labeling errors in auto-annotated images • Improved precision of bounding box localization • Enhanced classification granularity for multiple categories • Supported annotation best practices for industrial datasets

2030 - Present

outlier

CVATImagePrompt Response Writing SFTClassification

Worked on AI training and data labeling tasks involving prompt-response evaluation and content annotation. Reviewed AI-generated responses, ensured accuracy, and improved dataset quality for machine learning model training. Performed detailed analysis of image and text data, followed quality guidelines, and maintained consistency in labeling to support AI model development.

Worked on AI training and data labeling tasks involving prompt-response evaluation and content annotation. Reviewed AI-generated responses, ensured accuracy, and improved dataset quality for machine learning model training. Performed detailed analysis of image and text data, followed quality guidelines, and maintained consistency in labeling to support AI model development.

2025 - Present

AI Model Training and Evaluation Annotator

MindriftText

Evaluated and ranked AI-generated responses on Outlier AI and Mindrift platforms. Labeled textual data for contextual relevance, clarity, and linguistic quality. Provided feedback for improving large language model outputs. • Used structured rating systems to assess responses • Annotated data following precise guidelines • Identified areas for improvement in AI communication • Helped train conversational AI via quality metrics

Evaluated and ranked AI-generated responses on Outlier AI and Mindrift platforms. Labeled textual data for contextual relevance, clarity, and linguistic quality. Provided feedback for improving large language model outputs. • Used structured rating systems to assess responses • Annotated data following precise guidelines • Identified areas for improvement in AI communication • Helped train conversational AI via quality metrics

2023 - Present

AI Text Evaluation Practice Project

Text

Compared and ranked AI-generated textual outputs for empathy, clarity, and contextual alignment. Applied structured evaluation methods to ensure consistency and improve emotional tone. Rewrote model responses to enhance overall communication quality. • Utilized established scoring rubrics for evaluation • Improved AI-human communication fidelity • Provided actionable feedback on language models • Ensured emotional appropriateness in generated responses

Compared and ranked AI-generated textual outputs for empathy, clarity, and contextual alignment. Applied structured evaluation methods to ensure consistency and improve emotional tone. Rewrote model responses to enhance overall communication quality. • Utilized established scoring rubrics for evaluation • Improved AI-human communication fidelity • Provided actionable feedback on language models • Ensured emotional appropriateness in generated responses

2025 - 2025

LLM Response Evaluation Project

Text

Assessed and evaluated large language model (LLM) responses for clarity, empathy, and contextual accuracy. Compared and ranked outputs from multiple models, rewriting responses as needed for quality. Applied systematic, human-in-the-loop methods for AI text evaluation. • Used structured frameworks for AI response assessment • Conducted error analysis and quality improvement on LLMs • Enhanced model alignment through response rewriting • Provided valuable feedback for conversational AI models

Assessed and evaluated large language model (LLM) responses for clarity, empathy, and contextual accuracy. Compared and ranked outputs from multiple models, rewriting responses as needed for quality. Applied systematic, human-in-the-loop methods for AI text evaluation. • Used structured frameworks for AI response assessment • Conducted error analysis and quality improvement on LLMs • Enhanced model alignment through response rewriting • Provided valuable feedback for conversational AI models

2025 - 2025

Show more

Education

Work History